Package org.deeplearning4j.models.paragraphvectors

Class ParagraphVectors.Builder

java.lang.Object

org.deeplearning4j.models.sequencevectors.SequenceVectors.Builder<VocabWord>

org.deeplearning4j.models.word2vec.Word2Vec.Builder

org.deeplearning4j.models.paragraphvectors.ParagraphVectors.Builder

Enclosing class:

ParagraphVectors

public static class ParagraphVectors.Builder
extends Word2Vec.Builder

Field Summary

Fields

Modifier and Type	Field	Description
protected DocumentIterator	docIter
protected LabelAwareIterator	labelAwareIterator
protected LabelsSource	labelsSource

Fields inherited from class org.deeplearning4j.models.word2vec.Word2Vec.Builder

allowParallelTokenization, sentenceIterator, tokenizerFactory

Fields inherited from class org.deeplearning4j.models.sequencevectors.SequenceVectors.Builder

batchSize, configuration, elementsLearningAlgorithm, enableScavenger, existingVectors, hugeModelExpected, intersectVectors, iterations, iterator, layerSize, learningRate, learningRateDecayWords, lockFactor, lookupTable, minLearningRate, minWordFrequency, modelUtils, negative, numEpochs, preciseMode, preciseWeightInit, resetModel, sampling, seed, sequenceLearningAlgorithm, STOP, stopWords, trainElementsVectors, trainSequenceVectors, UNK, unknownElement, useAdaGrad, useHierarchicSoftmax, useUnknown, variableWindows, vectorsListeners, vocabCache, vocabLimit, window, workers

Constructor Summary

Constructors

Constructor	Description
Builder()
Builder(@NonNull VectorsConfiguration configuration)

Method Summary

Modifier and Type	Method	Description
ParagraphVectors.Builder	allowParallelTokenization(boolean allow)	This method enables/disables parallel tokenization.
ParagraphVectors.Builder	batchSize(int batchSize)	This method defines mini-batch size
ParagraphVectors	build()	Build SequenceVectors instance with defined settings/options
ParagraphVectors.Builder	elementsLearningAlgorithm(String algorithm)	* Sets specific LearningAlgorithm as Elements Learning Algorithm
ParagraphVectors.Builder	elementsLearningAlgorithm(ElementsLearningAlgorithm<VocabWord> algorithm)	* Sets specific LearningAlgorithm as Elements Learning Algorithm
ParagraphVectors.Builder	enableScavenger(boolean reallyEnable)	This method ebables/disables periodical vocab truncation during construction Default value: disabled
ParagraphVectors.Builder	epochs(int numEpochs)	This method defines number of epochs (iterations over whole training corpus) for training
ParagraphVectors.Builder	iterate(@NonNull SequenceIterator<VocabWord> iterator)	This method used to feed SequenceIterator, that contains training corpus, into ParagraphVectors
ParagraphVectors.Builder	iterate(@NonNull DocumentIterator iterator)	This method used to feed DocumentIterator, that contains training corpus, into ParagraphVectors
ParagraphVectors.Builder	iterate(@NonNull LabelAwareDocumentIterator iterator)	This method used to feed LabelAwareDocumentIterator, that contains training corpus, into ParagraphVectors
ParagraphVectors.Builder	iterate(@NonNull LabelAwareIterator iterator)	This method used to feed LabelAwareIterator, that contains training corpus, into ParagraphVectors
ParagraphVectors.Builder	iterate(@NonNull LabelAwareSentenceIterator iterator)	This method used to feed LabelAwareSentenceIterator, that contains training corpus, into ParagraphVectors
ParagraphVectors.Builder	iterate(@NonNull SentenceIterator iterator)	This method used to feed SentenceIterator, that contains training corpus, into ParagraphVectors
ParagraphVectors.Builder	iterations(int iterations)	This method defines number of iterations done for each mini-batch during training
ParagraphVectors.Builder	labels(@NonNull List<String> labels)	Deprecated.
ParagraphVectors.Builder	labelsSource(@NonNull LabelsSource source)	This method attaches pre-defined labels source to ParagraphVectors
ParagraphVectors.Builder	layerSize(int layerSize)	This method defines number of dimensions for output vectors
ParagraphVectors.Builder	learningRate(double learningRate)	This method defines initial learning rate for model training
ParagraphVectors.Builder	limitVocabularySize(int limit)	This method sets vocabulary limit during construction.
ParagraphVectors.Builder	lookupTable(@NonNull WeightLookupTable<VocabWord> lookupTable)	This method allows to define external WeightLookupTable to be used
ParagraphVectors.Builder	minLearningRate(double minLearningRate)	This method defines minimal learning rate value for training
ParagraphVectors.Builder	minWordFrequency(int minWordFrequency)	This method defines minimal word frequency in training corpus.
ParagraphVectors.Builder	modelUtils(@NonNull ModelUtils<VocabWord> modelUtils)	Sets ModelUtils that gonna be used as provider for utility methods: similarity(), wordsNearest(), accuracy(), etc
ParagraphVectors.Builder	negativeSample(double negative)	This method defines whether negative sampling should be used or not PLEASE NOTE: If you're going to use negative sampling, you might want to disable HierarchicSoftmax, which is enabled by default Default value: 0
ParagraphVectors.Builder	resetModel(boolean reallyReset)	This method defines whether model should be totally wiped out prior building, or not
ParagraphVectors.Builder	sampling(double sampling)	This method defines whether subsampling should be used or not
ParagraphVectors.Builder	seed(long randomSeed)	This method defines random seed for random numbers generator
ParagraphVectors.Builder	sequenceLearningAlgorithm(String algorithm)	Sets specific LearningAlgorithm as Sequence Learning Algorithm
ParagraphVectors.Builder	sequenceLearningAlgorithm(SequenceLearningAlgorithm<VocabWord> algorithm)	Sets specific LearningAlgorithm as Sequence Learning Algorithm
ParagraphVectors.Builder	setVectorsListeners(@NonNull Collection<VectorsListener<VocabWord>> vectorsListeners)	This method sets VectorsListeners for this SequenceVectors model
ParagraphVectors.Builder	stopWords(@NonNull Collection<VocabWord> stopList)	This method defines stop words that should be ignored during training
ParagraphVectors.Builder	stopWords(@NonNull List<String> stopList)	This method defines stop words that should be ignored during training
ParagraphVectors.Builder	tokenizerFactory(@NonNull TokenizerFactory tokenizerFactory)	This method defines TokenizerFactory to be used for strings tokenization during training PLEASE NOTE: If external VocabCache is used, the same TokenizerFactory should be used to keep derived tokens equal.
ParagraphVectors.Builder	trainElementsRepresentation(boolean trainElements)	This method defines, if words representation should be build together with documents representations.
ParagraphVectors.Builder	trainSequencesRepresentation(boolean trainSequences)	This method is hardcoded to TRUE, since that's whole point of ParagraphVectors
ParagraphVectors.Builder	trainWordVectors(boolean trainElements)	This method defines, if words representations should be build together with documents representations.
ParagraphVectors.Builder	unknownElement(VocabWord element)	This method allows you to specify SequenceElement that will be used as UNK element, if UNK is used
ParagraphVectors.Builder	useAdaGrad(boolean reallyUse)	This method defines whether adaptive gradients should be used or not
ParagraphVectors.Builder	useExistingWordVectors(@NonNull WordVectors vec)	This method allows you to use pre-built WordVectors model (e.g.
ParagraphVectors.Builder	useHierarchicSoftmax(boolean reallyUse)	This method enables/disables Hierarchic softmax Default value: enabled
ParagraphVectors.Builder	usePreciseWeightInit(boolean reallyUse)	If set to true, initial weights for elements/sequences will be derived from elements themself.
ParagraphVectors.Builder	useUnknown(boolean reallyUse)	This method allows you to specify, if UNK word should be used internally
ParagraphVectors.Builder	useVariableWindow(int... windows)	This method has no effect for ParagraphVectors
ParagraphVectors.Builder	vocabCache(@NonNull VocabCache<VocabWord> vocabCache)	This method allows to define external VocabCache to be used
ParagraphVectors.Builder	windowSize(int windowSize)	This method defines context window size
ParagraphVectors.Builder	workers(int numWorkers)	This method defines maximum number of concurrent threads available for training

Methods inherited from class org.deeplearning4j.models.word2vec.Word2Vec.Builder

intersectModel, usePreciseMode

Methods inherited from class org.deeplearning4j.models.sequencevectors.SequenceVectors.Builder

presetTables

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

labelAwareIterator

protected LabelAwareIterator labelAwareIterator

labelsSource

protected LabelsSource labelsSource

docIter

protected DocumentIterator docIter

Constructor Detail

Builder

public Builder()

Builder

public Builder(@NonNull
               @NonNull VectorsConfiguration configuration)

Method Detail

useExistingWordVectors

public ParagraphVectors.Builder useExistingWordVectors(@NonNull
                                                       @NonNull WordVectors vec)

This method allows you to use pre-built WordVectors model (e.g. Word2Vec) for ParagraphVectors. Existing model will be transferred into new model before training starts. PLEASE NOTE: Non-normalized model is recommended to use here.

Overrides:

useExistingWordVectors in class Word2Vec.Builder

Parameters:

vec - existing WordVectors model

Returns:

trainWordVectors

public ParagraphVectors.Builder trainWordVectors(boolean trainElements)

This method defines, if words representations should be build together with documents representations.

Parameters:

trainElements -

Returns:

labelsSource

public ParagraphVectors.Builder labelsSource(@NonNull
                                             @NonNull LabelsSource source)

This method attaches pre-defined labels source to ParagraphVectors

Parameters:

source -

Returns:

labels

@Deprecated
public ParagraphVectors.Builder labels(@NonNull
                                       @NonNull List<String> labels)

Deprecated.

This method builds new LabelSource instance from labels. PLEASE NOTE: Order synchro between labels and input documents delegated to end-user. PLEASE NOTE: Due to order issues it's recommended to use label aware iterators instead.

Parameters:

labels -

Returns:

iterate

public ParagraphVectors.Builder iterate(@NonNull
                                        @NonNull LabelAwareDocumentIterator iterator)

This method used to feed LabelAwareDocumentIterator, that contains training corpus, into ParagraphVectors

Parameters:

iterator -

Returns:

iterate

public ParagraphVectors.Builder iterate(@NonNull
                                        @NonNull LabelAwareSentenceIterator iterator)

This method used to feed LabelAwareSentenceIterator, that contains training corpus, into ParagraphVectors

Parameters:

iterator -

Returns:

iterate

public ParagraphVectors.Builder iterate(@NonNull
                                        @NonNull LabelAwareIterator iterator)

This method used to feed LabelAwareIterator, that contains training corpus, into ParagraphVectors

Overrides:

iterate in class Word2Vec.Builder

Parameters:

iterator -

Returns:

iterate

public ParagraphVectors.Builder iterate(@NonNull
                                        @NonNull DocumentIterator iterator)

This method used to feed DocumentIterator, that contains training corpus, into ParagraphVectors

Overrides:

iterate in class Word2Vec.Builder

Parameters:

iterator -

Returns:

iterate

public ParagraphVectors.Builder iterate(@NonNull
                                        @NonNull SentenceIterator iterator)

This method used to feed SentenceIterator, that contains training corpus, into ParagraphVectors

Overrides:

iterate in class Word2Vec.Builder

Parameters:

iterator -

Returns:

modelUtils

public ParagraphVectors.Builder modelUtils(@NonNull
                                           @NonNull ModelUtils<VocabWord> modelUtils)

Sets ModelUtils that gonna be used as provider for utility methods: similarity(), wordsNearest(), accuracy(), etc

Overrides:

modelUtils in class Word2Vec.Builder

Parameters:

modelUtils - model utils to be used

Returns:

limitVocabularySize

public ParagraphVectors.Builder limitVocabularySize(int limit)

This method sets vocabulary limit during construction. Default value: 0. Means no limit

Overrides:

limitVocabularySize in class Word2Vec.Builder

Parameters:

limit -

Returns:

unknownElement

public ParagraphVectors.Builder unknownElement(VocabWord element)

This method allows you to specify SequenceElement that will be used as UNK element, if UNK is used

Overrides:

unknownElement in class Word2Vec.Builder

Parameters:

element -

Returns:

allowParallelTokenization

public ParagraphVectors.Builder allowParallelTokenization(boolean allow)

This method enables/disables parallel tokenization. Default value: TRUE

Overrides:

allowParallelTokenization in class Word2Vec.Builder

Parameters:

allow -

Returns:

useUnknown

public ParagraphVectors.Builder useUnknown(boolean reallyUse)

This method allows you to specify, if UNK word should be used internally

Overrides:

useUnknown in class Word2Vec.Builder

Parameters:

reallyUse -

Returns:

enableScavenger

public ParagraphVectors.Builder enableScavenger(boolean reallyEnable)

This method ebables/disables periodical vocab truncation during construction Default value: disabled

Overrides:

enableScavenger in class Word2Vec.Builder

Parameters:

reallyEnable -

Returns:

build

public ParagraphVectors build()

Description copied from class: SequenceVectors.Builder

Build SequenceVectors instance with defined settings/options

Overrides:

build in class Word2Vec.Builder

Returns:

tokenizerFactory

public ParagraphVectors.Builder tokenizerFactory(@NonNull
                                                 @NonNull TokenizerFactory tokenizerFactory)

This method defines TokenizerFactory to be used for strings tokenization during training PLEASE NOTE: If external VocabCache is used, the same TokenizerFactory should be used to keep derived tokens equal.

Overrides:

tokenizerFactory in class Word2Vec.Builder

Parameters:

tokenizerFactory -

Returns:

iterate

public ParagraphVectors.Builder iterate(@NonNull
                                        @NonNull SequenceIterator<VocabWord> iterator)

This method used to feed SequenceIterator, that contains training corpus, into ParagraphVectors

Overrides:

iterate in class Word2Vec.Builder

Parameters:

iterator -

Returns:

batchSize

public ParagraphVectors.Builder batchSize(int batchSize)

This method defines mini-batch size

Overrides:

batchSize in class Word2Vec.Builder

Parameters:

batchSize -

Returns:

iterations

public ParagraphVectors.Builder iterations(int iterations)

This method defines number of iterations done for each mini-batch during training

Overrides:

iterations in class Word2Vec.Builder

Parameters:

iterations -

Returns:

epochs

public ParagraphVectors.Builder epochs(int numEpochs)

This method defines number of epochs (iterations over whole training corpus) for training

Overrides:

epochs in class Word2Vec.Builder

Parameters:

numEpochs -

Returns:

layerSize

public ParagraphVectors.Builder layerSize(int layerSize)

This method defines number of dimensions for output vectors

Overrides:

layerSize in class Word2Vec.Builder

Parameters:

layerSize -

Returns:

setVectorsListeners

public ParagraphVectors.Builder setVectorsListeners(@NonNull
                                                    @NonNull Collection<VectorsListener<VocabWord>> vectorsListeners)

This method sets VectorsListeners for this SequenceVectors model

Overrides:

setVectorsListeners in class Word2Vec.Builder

Parameters:

vectorsListeners -

Returns:

learningRate

public ParagraphVectors.Builder learningRate(double learningRate)

This method defines initial learning rate for model training

Overrides:

learningRate in class Word2Vec.Builder

Parameters:

learningRate -

Returns:

minWordFrequency

public ParagraphVectors.Builder minWordFrequency(int minWordFrequency)

This method defines minimal word frequency in training corpus. All words below this threshold will be removed prior model training

Overrides:

minWordFrequency in class Word2Vec.Builder

Parameters:

minWordFrequency -

Returns:

minLearningRate

public ParagraphVectors.Builder minLearningRate(double minLearningRate)

This method defines minimal learning rate value for training

Overrides:

minLearningRate in class Word2Vec.Builder

Parameters:

minLearningRate -

Returns:

resetModel

public ParagraphVectors.Builder resetModel(boolean reallyReset)

This method defines whether model should be totally wiped out prior building, or not

Overrides:

resetModel in class Word2Vec.Builder

Parameters:

reallyReset -

Returns:

vocabCache

public ParagraphVectors.Builder vocabCache(@NonNull
                                           @NonNull VocabCache<VocabWord> vocabCache)

This method allows to define external VocabCache to be used

Overrides:

vocabCache in class Word2Vec.Builder

Parameters:

vocabCache -

Returns:

lookupTable

public ParagraphVectors.Builder lookupTable(@NonNull
                                            @NonNull WeightLookupTable<VocabWord> lookupTable)

This method allows to define external WeightLookupTable to be used

Overrides:

lookupTable in class Word2Vec.Builder

Parameters:

lookupTable -

Returns:

sampling

public ParagraphVectors.Builder sampling(double sampling)

This method defines whether subsampling should be used or not

Overrides:

sampling in class Word2Vec.Builder

Parameters:

sampling - set > 0 to subsampling argument, or 0 to disable

Returns:

useAdaGrad

public ParagraphVectors.Builder useAdaGrad(boolean reallyUse)

This method defines whether adaptive gradients should be used or not

Overrides:

useAdaGrad in class Word2Vec.Builder

Parameters:

reallyUse -

Returns:

negativeSample

public ParagraphVectors.Builder negativeSample(double negative)

This method defines whether negative sampling should be used or not PLEASE NOTE: If you're going to use negative sampling, you might want to disable HierarchicSoftmax, which is enabled by default Default value: 0

Overrides:

negativeSample in class Word2Vec.Builder

Parameters:

negative - set > 0 as negative sampling argument, or 0 to disable

Returns:

stopWords

public ParagraphVectors.Builder stopWords(@NonNull
                                          @NonNull List<String> stopList)

This method defines stop words that should be ignored during training

Overrides:

stopWords in class Word2Vec.Builder

Parameters:

stopList -

Returns:

trainElementsRepresentation

public ParagraphVectors.Builder trainElementsRepresentation(boolean trainElements)

This method defines, if words representation should be build together with documents representations.

Overrides:

trainElementsRepresentation in class Word2Vec.Builder

Parameters:

trainElements -

Returns:

trainSequencesRepresentation

public ParagraphVectors.Builder trainSequencesRepresentation(boolean trainSequences)

This method is hardcoded to TRUE, since that's whole point of ParagraphVectors

Overrides:

trainSequencesRepresentation in class Word2Vec.Builder

Parameters:

trainSequences -

Returns:

stopWords

public ParagraphVectors.Builder stopWords(@NonNull
                                          @NonNull Collection<VocabWord> stopList)

This method defines stop words that should be ignored during training

Overrides:

stopWords in class Word2Vec.Builder

Parameters:

stopList -

Returns:

windowSize

public ParagraphVectors.Builder windowSize(int windowSize)

This method defines context window size

Overrides:

windowSize in class Word2Vec.Builder

Parameters:

windowSize -

Returns:

workers

public ParagraphVectors.Builder workers(int numWorkers)

This method defines maximum number of concurrent threads available for training

Overrides:

workers in class Word2Vec.Builder

Parameters:

numWorkers -

Returns:

sequenceLearningAlgorithm

public ParagraphVectors.Builder sequenceLearningAlgorithm(SequenceLearningAlgorithm<VocabWord> algorithm)

Description copied from class: SequenceVectors.Builder

Sets specific LearningAlgorithm as Sequence Learning Algorithm

Overrides:

sequenceLearningAlgorithm in class SequenceVectors.Builder<VocabWord>

Parameters:

algorithm - SequenceLearningAlgorithm implementation

Returns:

sequenceLearningAlgorithm

public ParagraphVectors.Builder sequenceLearningAlgorithm(String algorithm)

Description copied from class: SequenceVectors.Builder

Sets specific LearningAlgorithm as Sequence Learning Algorithm

Overrides:

sequenceLearningAlgorithm in class SequenceVectors.Builder<VocabWord>

Parameters:

algorithm - fully qualified class name

Returns:

useHierarchicSoftmax

public ParagraphVectors.Builder useHierarchicSoftmax(boolean reallyUse)

This method enables/disables Hierarchic softmax Default value: enabled

Overrides:

useHierarchicSoftmax in class Word2Vec.Builder

Parameters:

reallyUse -

Returns:

useVariableWindow

public ParagraphVectors.Builder useVariableWindow(int... windows)

This method has no effect for ParagraphVectors

Overrides:

useVariableWindow in class Word2Vec.Builder

Parameters:

windows -

Returns:

elementsLearningAlgorithm

public ParagraphVectors.Builder elementsLearningAlgorithm(ElementsLearningAlgorithm<VocabWord> algorithm)

Description copied from class: SequenceVectors.Builder

* Sets specific LearningAlgorithm as Elements Learning Algorithm

Overrides:

elementsLearningAlgorithm in class Word2Vec.Builder

Parameters:

algorithm - ElementsLearningAlgorithm implementation

Returns:

elementsLearningAlgorithm

public ParagraphVectors.Builder elementsLearningAlgorithm(String algorithm)

Description copied from class: SequenceVectors.Builder

* Sets specific LearningAlgorithm as Elements Learning Algorithm

Overrides:

elementsLearningAlgorithm in class Word2Vec.Builder

Parameters:

algorithm - fully qualified class name

Returns:

usePreciseWeightInit

public ParagraphVectors.Builder usePreciseWeightInit(boolean reallyUse)

Description copied from class: SequenceVectors.Builder

If set to true, initial weights for elements/sequences will be derived from elements themself. However, this implies additional cycle through input iterator. Default value: FALSE

Overrides:

usePreciseWeightInit in class Word2Vec.Builder

Returns:

seed

public ParagraphVectors.Builder seed(long randomSeed)

This method defines random seed for random numbers generator

Overrides:

seed in class Word2Vec.Builder

Parameters:

randomSeed -

Returns: